本文選題：智能視頻監(jiān)控 + HOFO��； 參考：《南京航空航天大學(xué)》2017年碩士論文

【摘要】：隨著中國民航業(yè)的迅猛發(fā)展,機(jī)場(chǎng)所面臨的安全壓力日益增大。本文針對(duì)傳統(tǒng)視頻監(jiān)控系統(tǒng)局限于人力監(jiān)視缺乏主動(dòng)檢測(cè)識(shí)別異常事件的能力,對(duì)智能視頻監(jiān)控系統(tǒng)(Intelligent Video Surveillance,簡稱IVS)中的異常行為檢測(cè)算法與目標(biāo)跟蹤算法進(jìn)行了研究,旨在為高性能的智能視頻監(jiān)控系統(tǒng)的開發(fā)與實(shí)現(xiàn)提供思路和參考。本文主要研究內(nèi)容如下:針對(duì)異常行為的檢測(cè)問題,本文對(duì)傳統(tǒng)的基于光流方向直方圖(Histogram of Optical Flow Orientation,簡稱HOFO)的異常檢測(cè)方法進(jìn)行了改進(jìn)。傳統(tǒng)的光流方向直方圖的計(jì)算僅限于對(duì)光流方向的簡單統(tǒng)計(jì),對(duì)圖像信息的描述存在不足,為了提高對(duì)圖像信息的表達(dá)能力,本文中對(duì)直方圖的統(tǒng)計(jì)方式進(jìn)行了改進(jìn),將光流能量加權(quán)到直方圖的計(jì)算中,提出了一種基于加權(quán)光流能量的HOFO特征的異常行為檢測(cè)算法。實(shí)驗(yàn)表明,改進(jìn)后的算法與原始算法相比檢測(cè)準(zhǔn)確率得到了一定程度的提高。針對(duì)上述異常檢測(cè)算法檢測(cè)速度與準(zhǔn)確率低的問題,本文將基于卷積神經(jīng)網(wǎng)絡(luò)(Convolutional Neural Networks,簡稱CNN)應(yīng)用于異常行為的檢測(cè),提出了一種基于卷積神經(jīng)網(wǎng)絡(luò)的異常行為檢測(cè)算法。該算法不需要設(shè)計(jì)特征提取器,可以直接將圖像作為輸入,同時(shí)又采用了局部感知和權(quán)值共享的方法,大大加快了算法速度。實(shí)驗(yàn)表明,該算法相對(duì)于上述異常行為檢測(cè)算法不僅加快了算法速度而且提高了檢測(cè)準(zhǔn)確率。針對(duì)現(xiàn)有跟蹤算法遇到遮擋、形變、以及光照變化而引起的跟蹤失敗問題,本文提出了一種融合表觀特征與深度特征的目標(biāo)跟蹤算法。首先用大量行人數(shù)據(jù)庫對(duì)CNN網(wǎng)絡(luò)進(jìn)行訓(xùn)練,然后用訓(xùn)練好的CNN網(wǎng)絡(luò)提取目標(biāo)區(qū)域的深度特征,同時(shí)計(jì)算目標(biāo)區(qū)域在HSV空間的顏色直方圖,將深度特征與顏色特征進(jìn)行聯(lián)合得到整體特征。最后在粒子濾波框架下對(duì)多個(gè)假設(shè)狀態(tài)進(jìn)行估計(jì),獲得目標(biāo)的最優(yōu)位置,得到跟蹤結(jié)果,并進(jìn)行模板更新,最后根據(jù)粒子的退化情況,進(jìn)行重采樣。實(shí)驗(yàn)表明,本文跟蹤算法獲得了良好的跟蹤魯棒性。最后,設(shè)計(jì)了異常行為檢測(cè)與目標(biāo)跟蹤系統(tǒng)并在Matlab平臺(tái)上進(jìn)行了仿真實(shí)現(xiàn),驗(yàn)證了本論文所研究算法的有效性和實(shí)用性。
[Abstract]:With the rapid development of China's civil aviation industry, the airport is facing increasing safety pressure. In this paper, the traditional video surveillance system is limited to human monitoring, which lacks the ability to detect and identify abnormal events. In this paper, the detection algorithm of abnormal behavior and the algorithm of target tracking in intelligent video surveillance system (Intelligent Video Survey) are studied. The aim is to provide ideas and references for the development and implementation of intelligent video surveillance system with high performance. The main contents of this paper are as follows: aiming at the problem of abnormal behavior detection, this paper improves the traditional anomaly detection method based on histogram (HOFOO). The traditional calculation of optical flow direction histogram is limited to the simple statistics of the optical flow direction, but the description of the image information is insufficient. In order to improve the expression ability of the image information, the statistical method of the histogram is improved in this paper. In this paper, the optical flow energy is weighted to the histogram, and an anomaly detection algorithm based on the weighted optical flow energy HOFO feature is proposed. Experiments show that the detection accuracy of the improved algorithm is improved to some extent compared with the original algorithm. In order to solve the problem of low detection speed and accuracy, this paper presents an anomaly detection algorithm based on convolution neural network (CNNs) based on Convolutional Neural Networks (CNNs). The algorithm does not need to design a feature extractor and can directly use the image as the input. At the same time, it uses the method of local perception and weight sharing, which greatly accelerates the speed of the algorithm. Experimental results show that the proposed algorithm not only speeds up the speed of the algorithm but also improves the detection accuracy compared with the above algorithms. In order to solve the problem of tracking failure caused by occlusion, deformation and illumination change, a target tracking algorithm combining apparent features and depth features is proposed in this paper. Firstly, the CNN network is trained with a large number of pedestrian databases, then the depth features of the target area are extracted by the trained CNN network, and the color histogram of the target area in the HSV space is calculated at the same time. The depth feature is combined with the color feature to get the whole feature. Finally, several hypothetical states are estimated in the framework of particle filter, the optimal location of the target is obtained, the tracking results are obtained, and the template is updated. Finally, according to the degradation of particles, the sample is re-sampled. Experiments show that the proposed tracking algorithm has good tracking robustness. Finally, the anomaly behavior detection and target tracking system is designed and implemented on Matlab platform, which verifies the validity and practicability of the algorithm studied in this paper.
【學(xué)位授予單位】：南京航空航天大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：TN948.6

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李寰宇;畢篤彥;楊源;查宇飛;覃兵;張立朝;;基于深度特征表達(dá)與學(xué)習(xí)的視覺跟蹤算法研究[J];電子與信息學(xué)報(bào);2015年09期

2 傅博;李文輝;陳博;王聰;王瑩;;基于加權(quán)光流能量的異常行為檢測(cè)[J];吉林大學(xué)學(xué)報(bào)(工學(xué)版);2013年06期

3 謝天旭;何炳蔚;;基于顏色特征與SIFT特征自適應(yīng)融合的粒子濾波跟蹤算法[J];計(jì)算機(jī)應(yīng)用研究;2013年11期

4 王煒;郭毓;俞信;;基于卡爾曼濾波的多區(qū)域關(guān)聯(lián)運(yùn)動(dòng)目標(biāo)跟蹤[J];計(jì)算機(jī)應(yīng)用;2012年11期

5 劉連忠;寧海燕;;基于支持向量機(jī)的車牌字符識(shí)別[J];微計(jì)算機(jī)信息;2012年10期

6 楚瀛;符楊;;馬爾可夫隨機(jī)場(chǎng)理論在目標(biāo)跟蹤中的應(yīng)用[J];計(jì)算機(jī)工程與應(yīng)用;2011年28期

7 劉治紅;駱云志;;智能視頻監(jiān)控技術(shù)在哨位安全威脅智能識(shí)別中的應(yīng)用[J];兵工自動(dòng)化;2011年07期

8 彭欣;劉富強(qiáng);宋華軍;;基于SURF目標(biāo)跟蹤算法研究[J];長春理工大學(xué)學(xué)報(bào)(自然科學(xué)版);2011年02期

9 鄭楚容;王東偉;洪天jR;蔣敏捷;龐國明;;民航安防市場(chǎng)迎來高速發(fā)展期[J];中國安防;2010年10期

10 藺海峰;馬宇峰;宋濤;;基于SIFT特征目標(biāo)跟蹤算法研究[J];自動(dòng)化學(xué)報(bào);2010年08期

相關(guān)博士學(xué)位論文 前2條

1 李彤;智能視頻監(jiān)控下的多目標(biāo)跟蹤技術(shù)研究[D];中國科學(xué)技術(shù)大學(xué);2013年

2 鄭艷;MPEG-2到H.264/AVC視頻轉(zhuǎn)碼及相關(guān)技術(shù)研究[D];上海交通大學(xué);2008年

相關(guān)碩士學(xué)位論文 前3條

1 張子夫;基于卷積神經(jīng)網(wǎng)絡(luò)的目標(biāo)跟蹤算法研究與實(shí)現(xiàn)[D];吉林大學(xué);2015年

2 劉潔;基于光流法的運(yùn)動(dòng)目標(biāo)檢測(cè)和跟蹤算法研究[D];中國礦業(yè)大學(xué);2015年

3 劉振;基于卡爾曼濾波與模板匹配的目標(biāo)跟蹤研究[D];昆明理工大學(xué);2011年

，

本文編號(hào)：1969443